{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.979036827195468,
  "eval_steps": 500,
  "global_step": 4410,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11331444759206799,
      "grad_norm": 0.12750307449685053,
      "learning_rate": 1.977324263038549e-05,
      "loss": 0.2329,
      "step": 50
    },
    {
      "epoch": 0.22662889518413598,
      "grad_norm": 0.10813549127897755,
      "learning_rate": 1.9546485260770977e-05,
      "loss": 0.0091,
      "step": 100
    },
    {
      "epoch": 0.33994334277620397,
      "grad_norm": 0.06931579383532684,
      "learning_rate": 1.9319727891156463e-05,
      "loss": 0.0118,
      "step": 150
    },
    {
      "epoch": 0.45325779036827196,
      "grad_norm": 0.05818309038266094,
      "learning_rate": 1.9092970521541953e-05,
      "loss": 0.004,
      "step": 200
    },
    {
      "epoch": 0.56657223796034,
      "grad_norm": 0.04003839601543832,
      "learning_rate": 1.886621315192744e-05,
      "loss": 0.003,
      "step": 250
    },
    {
      "epoch": 0.6798866855524079,
      "grad_norm": 0.05009081423940936,
      "learning_rate": 1.863945578231293e-05,
      "loss": 0.0025,
      "step": 300
    },
    {
      "epoch": 0.7932011331444759,
      "grad_norm": 0.043467219593294984,
      "learning_rate": 1.8412698412698415e-05,
      "loss": 0.0022,
      "step": 350
    },
    {
      "epoch": 0.9065155807365439,
      "grad_norm": 0.03628971874099008,
      "learning_rate": 1.81859410430839e-05,
      "loss": 0.002,
      "step": 400
    },
    {
      "epoch": 1.0181303116147309,
      "grad_norm": 0.03627715893292353,
      "learning_rate": 1.795918367346939e-05,
      "loss": 0.0017,
      "step": 450
    },
    {
      "epoch": 1.1314447592067989,
      "grad_norm": 0.032502169772711025,
      "learning_rate": 1.7732426303854877e-05,
      "loss": 0.0015,
      "step": 500
    },
    {
      "epoch": 1.2447592067988669,
      "grad_norm": 0.03800243357726477,
      "learning_rate": 1.7505668934240366e-05,
      "loss": 0.0015,
      "step": 550
    },
    {
      "epoch": 1.3580736543909349,
      "grad_norm": 0.03683426555432635,
      "learning_rate": 1.7278911564625852e-05,
      "loss": 0.0014,
      "step": 600
    },
    {
      "epoch": 1.4713881019830028,
      "grad_norm": 0.030197812160831627,
      "learning_rate": 1.705215419501134e-05,
      "loss": 0.0013,
      "step": 650
    },
    {
      "epoch": 1.5847025495750708,
      "grad_norm": 0.029519742432872642,
      "learning_rate": 1.6825396825396828e-05,
      "loss": 0.0013,
      "step": 700
    },
    {
      "epoch": 1.6980169971671388,
      "grad_norm": 0.031772318618878956,
      "learning_rate": 1.6598639455782314e-05,
      "loss": 0.0012,
      "step": 750
    },
    {
      "epoch": 1.8113314447592068,
      "grad_norm": 0.029907198061296256,
      "learning_rate": 1.63718820861678e-05,
      "loss": 0.0012,
      "step": 800
    },
    {
      "epoch": 1.9246458923512748,
      "grad_norm": 0.0470370171523219,
      "learning_rate": 1.614512471655329e-05,
      "loss": 0.0013,
      "step": 850
    },
    {
      "epoch": 2.0362606232294618,
      "grad_norm": 0.029126616132443304,
      "learning_rate": 1.5918367346938776e-05,
      "loss": 0.0011,
      "step": 900
    },
    {
      "epoch": 2.1495750708215295,
      "grad_norm": 0.025826203236221357,
      "learning_rate": 1.5691609977324265e-05,
      "loss": 0.0009,
      "step": 950
    },
    {
      "epoch": 2.2628895184135978,
      "grad_norm": 0.026184094780744977,
      "learning_rate": 1.546485260770975e-05,
      "loss": 0.001,
      "step": 1000
    },
    {
      "epoch": 2.376203966005666,
      "grad_norm": 0.02104685350387791,
      "learning_rate": 1.523809523809524e-05,
      "loss": 0.0009,
      "step": 1050
    },
    {
      "epoch": 2.4895184135977337,
      "grad_norm": 0.026962231904391707,
      "learning_rate": 1.5011337868480727e-05,
      "loss": 0.0009,
      "step": 1100
    },
    {
      "epoch": 2.6028328611898015,
      "grad_norm": 0.02386983035895265,
      "learning_rate": 1.4784580498866215e-05,
      "loss": 0.0009,
      "step": 1150
    },
    {
      "epoch": 2.7161473087818697,
      "grad_norm": 0.03135625507160954,
      "learning_rate": 1.4557823129251703e-05,
      "loss": 0.0009,
      "step": 1200
    },
    {
      "epoch": 2.829461756373938,
      "grad_norm": 0.021990829289825996,
      "learning_rate": 1.433106575963719e-05,
      "loss": 0.0008,
      "step": 1250
    },
    {
      "epoch": 2.9427762039660057,
      "grad_norm": 0.027975048412071702,
      "learning_rate": 1.4104308390022677e-05,
      "loss": 0.0008,
      "step": 1300
    },
    {
      "epoch": 3.0543909348441924,
      "grad_norm": 0.02125910446859899,
      "learning_rate": 1.3877551020408165e-05,
      "loss": 0.0008,
      "step": 1350
    },
    {
      "epoch": 3.1677053824362607,
      "grad_norm": 0.028731963252144437,
      "learning_rate": 1.3650793650793652e-05,
      "loss": 0.0007,
      "step": 1400
    },
    {
      "epoch": 3.2810198300283284,
      "grad_norm": 0.022590926398634654,
      "learning_rate": 1.342403628117914e-05,
      "loss": 0.0007,
      "step": 1450
    },
    {
      "epoch": 3.3943342776203966,
      "grad_norm": 0.021593995910696728,
      "learning_rate": 1.3197278911564626e-05,
      "loss": 0.0006,
      "step": 1500
    },
    {
      "epoch": 3.507648725212465,
      "grad_norm": 0.02666309225386757,
      "learning_rate": 1.2970521541950114e-05,
      "loss": 0.0006,
      "step": 1550
    },
    {
      "epoch": 3.6209631728045326,
      "grad_norm": 0.020932694755938938,
      "learning_rate": 1.2743764172335602e-05,
      "loss": 0.0006,
      "step": 1600
    },
    {
      "epoch": 3.7342776203966004,
      "grad_norm": 0.015749742123260534,
      "learning_rate": 1.251700680272109e-05,
      "loss": 0.0006,
      "step": 1650
    },
    {
      "epoch": 3.8475920679886686,
      "grad_norm": 0.02090008773458773,
      "learning_rate": 1.2290249433106578e-05,
      "loss": 0.0006,
      "step": 1700
    },
    {
      "epoch": 3.960906515580737,
      "grad_norm": 0.01537319079431243,
      "learning_rate": 1.2063492063492064e-05,
      "loss": 0.0006,
      "step": 1750
    },
    {
      "epoch": 4.0725212464589235,
      "grad_norm": 0.016456680181818843,
      "learning_rate": 1.1836734693877552e-05,
      "loss": 0.0005,
      "step": 1800
    },
    {
      "epoch": 4.185835694050992,
      "grad_norm": 0.014279201845786642,
      "learning_rate": 1.160997732426304e-05,
      "loss": 0.0004,
      "step": 1850
    },
    {
      "epoch": 4.299150141643059,
      "grad_norm": 0.013506804613582977,
      "learning_rate": 1.1383219954648527e-05,
      "loss": 0.0004,
      "step": 1900
    },
    {
      "epoch": 4.412464589235127,
      "grad_norm": 0.017575682502776736,
      "learning_rate": 1.1156462585034013e-05,
      "loss": 0.0004,
      "step": 1950
    },
    {
      "epoch": 4.5257790368271955,
      "grad_norm": 0.020440730932576385,
      "learning_rate": 1.0929705215419501e-05,
      "loss": 0.0004,
      "step": 2000
    },
    {
      "epoch": 4.639093484419264,
      "grad_norm": 0.01834736863465592,
      "learning_rate": 1.0702947845804989e-05,
      "loss": 0.0004,
      "step": 2050
    },
    {
      "epoch": 4.752407932011332,
      "grad_norm": 0.017860961767452527,
      "learning_rate": 1.0476190476190477e-05,
      "loss": 0.0004,
      "step": 2100
    },
    {
      "epoch": 4.865722379603399,
      "grad_norm": 0.020594689995493427,
      "learning_rate": 1.0249433106575966e-05,
      "loss": 0.0004,
      "step": 2150
    },
    {
      "epoch": 4.9790368271954675,
      "grad_norm": 0.019468621367640594,
      "learning_rate": 1.0022675736961451e-05,
      "loss": 0.0004,
      "step": 2200
    },
    {
      "epoch": 5.090651558073654,
      "grad_norm": 0.01815010750870599,
      "learning_rate": 9.795918367346939e-06,
      "loss": 0.0003,
      "step": 2250
    },
    {
      "epoch": 5.203966005665722,
      "grad_norm": 0.013915876229594861,
      "learning_rate": 9.569160997732427e-06,
      "loss": 0.0003,
      "step": 2300
    },
    {
      "epoch": 5.317280453257791,
      "grad_norm": 0.01012229778431893,
      "learning_rate": 9.342403628117914e-06,
      "loss": 0.0003,
      "step": 2350
    },
    {
      "epoch": 5.430594900849858,
      "grad_norm": 0.014316093650605412,
      "learning_rate": 9.115646258503402e-06,
      "loss": 0.0003,
      "step": 2400
    },
    {
      "epoch": 5.543909348441926,
      "grad_norm": 0.017567253599829386,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.0003,
      "step": 2450
    },
    {
      "epoch": 5.657223796033994,
      "grad_norm": 0.0226569072606562,
      "learning_rate": 8.662131519274378e-06,
      "loss": 0.0003,
      "step": 2500
    },
    {
      "epoch": 5.770538243626063,
      "grad_norm": 0.01682883062787434,
      "learning_rate": 8.435374149659866e-06,
      "loss": 0.0003,
      "step": 2550
    },
    {
      "epoch": 5.88385269121813,
      "grad_norm": 0.013391313567851492,
      "learning_rate": 8.208616780045352e-06,
      "loss": 0.0003,
      "step": 2600
    },
    {
      "epoch": 5.997167138810198,
      "grad_norm": 0.01778560595671548,
      "learning_rate": 7.98185941043084e-06,
      "loss": 0.0003,
      "step": 2650
    },
    {
      "epoch": 6.108781869688385,
      "grad_norm": 0.013117817816572433,
      "learning_rate": 7.755102040816327e-06,
      "loss": 0.0002,
      "step": 2700
    },
    {
      "epoch": 6.222096317280453,
      "grad_norm": 0.017618823564966208,
      "learning_rate": 7.528344671201815e-06,
      "loss": 0.0002,
      "step": 2750
    },
    {
      "epoch": 6.335410764872521,
      "grad_norm": 0.0096548743047912,
      "learning_rate": 7.301587301587301e-06,
      "loss": 0.0002,
      "step": 2800
    },
    {
      "epoch": 6.4487252124645895,
      "grad_norm": 0.010097830595674354,
      "learning_rate": 7.07482993197279e-06,
      "loss": 0.0002,
      "step": 2850
    },
    {
      "epoch": 6.562039660056657,
      "grad_norm": 0.010910197208914625,
      "learning_rate": 6.848072562358277e-06,
      "loss": 0.0002,
      "step": 2900
    },
    {
      "epoch": 6.675354107648725,
      "grad_norm": 0.012241078806903706,
      "learning_rate": 6.621315192743765e-06,
      "loss": 0.0002,
      "step": 2950
    },
    {
      "epoch": 6.788668555240793,
      "grad_norm": 0.017003802621226093,
      "learning_rate": 6.394557823129253e-06,
      "loss": 0.0002,
      "step": 3000
    },
    {
      "epoch": 6.9019830028328615,
      "grad_norm": 0.014847405573922268,
      "learning_rate": 6.16780045351474e-06,
      "loss": 0.0002,
      "step": 3050
    },
    {
      "epoch": 7.013597733711048,
      "grad_norm": 0.018305480145040905,
      "learning_rate": 5.9410430839002275e-06,
      "loss": 0.0002,
      "step": 3100
    },
    {
      "epoch": 7.126912181303116,
      "grad_norm": 0.012819862819578573,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.0001,
      "step": 3150
    },
    {
      "epoch": 7.240226628895184,
      "grad_norm": 0.013371528432039266,
      "learning_rate": 5.487528344671202e-06,
      "loss": 0.0001,
      "step": 3200
    },
    {
      "epoch": 7.353541076487252,
      "grad_norm": 0.005633164169599245,
      "learning_rate": 5.260770975056689e-06,
      "loss": 0.0001,
      "step": 3250
    },
    {
      "epoch": 7.46685552407932,
      "grad_norm": 0.01061442292135964,
      "learning_rate": 5.034013605442177e-06,
      "loss": 0.0001,
      "step": 3300
    },
    {
      "epoch": 7.580169971671388,
      "grad_norm": 0.00907172882592498,
      "learning_rate": 4.807256235827665e-06,
      "loss": 0.0001,
      "step": 3350
    },
    {
      "epoch": 7.693484419263456,
      "grad_norm": 0.010227726380358893,
      "learning_rate": 4.580498866213152e-06,
      "loss": 0.0001,
      "step": 3400
    },
    {
      "epoch": 7.806798866855524,
      "grad_norm": 0.008230239860332859,
      "learning_rate": 4.35374149659864e-06,
      "loss": 0.0001,
      "step": 3450
    },
    {
      "epoch": 7.920113314447592,
      "grad_norm": 0.013256768674645692,
      "learning_rate": 4.126984126984127e-06,
      "loss": 0.0001,
      "step": 3500
    },
    {
      "epoch": 8.03172804532578,
      "grad_norm": 0.01006097891003743,
      "learning_rate": 3.9002267573696154e-06,
      "loss": 0.0001,
      "step": 3550
    },
    {
      "epoch": 8.145042492917847,
      "grad_norm": 0.009602466515273548,
      "learning_rate": 3.6734693877551024e-06,
      "loss": 0.0001,
      "step": 3600
    },
    {
      "epoch": 8.258356940509914,
      "grad_norm": 0.005795654820959573,
      "learning_rate": 3.44671201814059e-06,
      "loss": 0.0001,
      "step": 3650
    },
    {
      "epoch": 8.371671388101984,
      "grad_norm": 0.0034753960503702283,
      "learning_rate": 3.2199546485260772e-06,
      "loss": 0.0001,
      "step": 3700
    },
    {
      "epoch": 8.48498583569405,
      "grad_norm": 0.006716712372637811,
      "learning_rate": 2.993197278911565e-06,
      "loss": 0.0001,
      "step": 3750
    },
    {
      "epoch": 8.598300283286118,
      "grad_norm": 0.005691017397956627,
      "learning_rate": 2.7664399092970525e-06,
      "loss": 0.0001,
      "step": 3800
    },
    {
      "epoch": 8.711614730878187,
      "grad_norm": 0.005512521725963195,
      "learning_rate": 2.53968253968254e-06,
      "loss": 0.0001,
      "step": 3850
    },
    {
      "epoch": 8.824929178470255,
      "grad_norm": 0.004574835289146862,
      "learning_rate": 2.3129251700680273e-06,
      "loss": 0.0001,
      "step": 3900
    },
    {
      "epoch": 8.938243626062324,
      "grad_norm": 0.005810520530240234,
      "learning_rate": 2.086167800453515e-06,
      "loss": 0.0001,
      "step": 3950
    },
    {
      "epoch": 9.04985835694051,
      "grad_norm": 0.0032799015697558206,
      "learning_rate": 1.8594104308390023e-06,
      "loss": 0.0001,
      "step": 4000
    },
    {
      "epoch": 9.163172804532579,
      "grad_norm": 0.0021797625948262924,
      "learning_rate": 1.6326530612244897e-06,
      "loss": 0.0,
      "step": 4050
    },
    {
      "epoch": 9.276487252124646,
      "grad_norm": 0.003847390080167819,
      "learning_rate": 1.4058956916099775e-06,
      "loss": 0.0,
      "step": 4100
    },
    {
      "epoch": 9.389801699716713,
      "grad_norm": 0.0025440319025627875,
      "learning_rate": 1.179138321995465e-06,
      "loss": 0.0,
      "step": 4150
    },
    {
      "epoch": 9.503116147308782,
      "grad_norm": 0.002651261609298388,
      "learning_rate": 9.523809523809525e-07,
      "loss": 0.0,
      "step": 4200
    },
    {
      "epoch": 9.61643059490085,
      "grad_norm": 0.0033509868369888817,
      "learning_rate": 7.2562358276644e-07,
      "loss": 0.0,
      "step": 4250
    },
    {
      "epoch": 9.729745042492917,
      "grad_norm": 0.011727033108113469,
      "learning_rate": 4.988662131519275e-07,
      "loss": 0.0,
      "step": 4300
    },
    {
      "epoch": 9.843059490084986,
      "grad_norm": 0.005342769601421854,
      "learning_rate": 2.72108843537415e-07,
      "loss": 0.0,
      "step": 4350
    },
    {
      "epoch": 9.956373937677053,
      "grad_norm": 0.004383322417463679,
      "learning_rate": 4.53514739229025e-08,
      "loss": 0.0,
      "step": 4400
    }
  ],
  "logging_steps": 50,
  "max_steps": 4410,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.446988850023629e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}